Abstract: An apparatus and method for detecting low level bioelectric signals on the surface of a patient. The bioelectric signals are enhanced by filtering out noise through a particular adaptive filtering technique. In detecting His signals at the chest of a patient, surface ECG's are acquired at a plurality of external locations, the acquired data is digitized and stored. One of the ECG signals is selected as a reference channel and the remaining ECG signals are used as input channels. With the use of a feedback coefficient, each data point of interest in each cycle is adaptively filtered to remove the noise and the filtered signal is displayed. The adaptive filtering involves performing two pre-updates of a weight matrix associated with each sample point of each input channel based on the measured samples before and after a sample point, calculating an output signal based on the weight matrix and the input channel, and updating the weight matrices based on the calculated output signal.